Many magnetic resonance (MR) modalities, like functional MR imaging (MRI), diffusion MRI and MR spectroscopy (MRS) have great potential for the study and diagnosis of brain disease and injury, and guiding surgical therapy. All of these methods benefit greatly from the increased sensitivity, resolution and contrast from high field strength magnets (3T and above). However, the advantages of higher magnetic fields have not been fully realized due to the increasingly confounding effects of magnetic field inhomogeneity caused by magnetic susceptibility differences between air and tissue. Magnetic field inhomogeneity leads to signal loss and spatial distortion in MRI and loss in spectral resolution and sensitivity in MRS. The loss of reliability due to these artifacts is a major reason why these techniques have not seen widespread use in high-field clinical applications. Current methods of magnetic field homogenization based on spherical harmonic shim coils perform well on small volumes but are inadequate to compensate the complex magnetic field distribution across the entire human brain. Here a novel technique based on matrices of DC coils is presented that can achieve superior shimming performance in the human brain in vivo. Preliminary results demonstrate the feasibility of multi-coil (MC) matrix shimming on human brain at 7.0 T. Characterization of the static and temporal characteristics of MC matrices will be established, as well as the integration with RF coils. This application proposes to develop MC matrix shimming to the point where excellent magnetic field homogeneity across the entire human brain at 7 T can be obtained in a robust and automated fashion for use with any existing clinical MR protocol. Specifically, the application is composed of three aims that are focused on optimization of the MC matrix for human brain applications at 7 T (Aim 1), the automation of MC matrix field generation (Aim 2) and demonstration of improved magnetic field homogeneity in a number of MRI, MRS and MRSI applications (Aim 3).